I'm checking use of SIMSOLID for random dynamics, e.g. for a quick assessment of local accelerations.
Due to the fact that SIMSOLID currently has only displacements PSD in output, I use "Omega arithmetic" to calculate accelerations, i.e. I multiply the calculated displacements PSD by omega4 to get the acceleration PSD.
As an usual check, I verify if the PSD acceleration calculated at constrained interfaces is the very same as the input PSD (normal work by Nastran / Optistruct).
Unfortunately results are non clear, as anyway I see an output that resembles the output in a generic point, even if I picked a point in a interface hole.
In the attachment, you can find a first sheet with an example of omega arithmetic by Nastran (i.e. conventional FEM) and everything is OK: despite units are a bit tricky, resulting PSD shape and levels are ok.
In the same file the second sheet reports the results from SIMSOLID, but neither the numbers nor the shape are ok.
Any feedback is appreciated.